Need faster way to get going involving Replication

[PDinCA]

Referencing post: http://www.red-gate.com/messageboard/viewtopic.php?t=12209

I also have issues bringing the core system down for 'n' hours while I:

Kill data acquisition
Kill Replication
Backup the Publisher DB (160GB+ data file)
Backup the Subscriber DB (35GB+ data file)
Restore Pub
Restore Sub
Rebuild Replication
(and that's if I don't do the recommended parallel database method.)

Based on current times, installing and commissioning on the Production server, SS2005 Enterprise, will take at least 6 hours, as rebuilding replication takes about 4 hours!

This seems like an entirely unreasonable length of time to take the core business system down - assuming all goes well end-to-end.

Is there no quicker way the transition can be effected? Are efforts underway down-under to perhaps make the downtime much shorter?

Those of us with large-enough DBs and Replication implications are presented with a hard pill to swallow, let alone convincing Management that the downtime window is "this big" and there's no other path to storage compression but through that hole...

Thoughts? Plans?

[eddie davis]

Hi, thank you your post into the forum.

Please accept my apologies, I am little confused my your post.

Are you seeking to change how SQL Storage Compress operates when you initially configure or setup a database to use the product?

Many Thanks
Eddie
_________________
Eddie Davis
Technical Support Engineer
Red Gate Software Ltd
E-mail: support@red-gate.com

[PDinCA]

Hi Eddie,

I'm having a very helpful conversation with Jeff @ HyperBac, courtesy of Angus. There is another way to enroll databases in SC but the downside appears to be that unless the data are very well physically organized at the point of enrollment, much lower compression ratios will be realized.

My problem lies in the fact that the 2 DBs I have are 160GB and 35GB data, with the 35GB being the subscriber to a single 124GB table in the larger DB. It takes about four hours to reestablish replication, so I was inquiring whether there was a way to minimize outage by not having to backup and restore each DB, then redo replaication from scratch.

We're nearing the end of the question and answer. I'll post what we find when complete.